IR2153 ETD49 700W SMPS problem

uxavier

New member
hello guys, im new to the forum i take the moment to thank the administrators for approval :)
this is my first approach to smps power supply design and building , for my first smps i decided to use ir2153 as is an oscillator with gate driver for Ho Lo integrated in one small package.
i studied all of his characteristics such as (Tr-Tf )(Sink-SourceCurrent) fixed duty cycle at 50% also dead time 1.2uS etc etc
i came up to the conclusion that with gate resistor ir2153 is not able to drive mosfet that have a Qg greater than 100ns at most
so i went with the approach of using a transistor for each mosfet instead to reduce false turn on caused by slow discharge of the mosfets
this is the schematic in wich i omitted:
  1. the soft start(relay with 200ms timer that bypasses 25ohm resistor on drain of HO mos)
  2. the emi filter with 15ohm Ntc at the input (double common mode choke with 2 1uf caps) .
  3. snubber parallel to the transformer (100ohm 1nf)
the transformer is an etd49 pc40TDK on wich i wound 25turns of 3x0.71 wire for the prim and 20turns of 3x0.71 wire (is the same wire size because i wanted to test)
those are the waveforms im getting out of this thing and im not happy at all, searching online and in various forums i kind of understand that every one that puts 2 mosfets along with a driver and crosses 2 wires to make a spark then proceeds to publish the schematic as fully working and *very good* 🤣
but basically i understand that if the inductance of the primary is too high this kind of phenomenon tends to occur (or at least this is what i think)
on the other hand in other forums people tell the opposite in wich if the inductance of the primary is too low this phenomenon occurs.
hands down im kinda flabbergasted about that.... so many wrong people and so many blown circuit breakers :poop:

i calculated my transformer using the formula provided in tahmid blog using a value of bmax of 0.18Tesla or 1800Gauss wich corresponded to 24.5turns (25for convenience)
the operating frequency of the circuit is 45khz, no matter what after 500w of load my mosfets start to heat abnormally (they are mounted on the same heatsink i cant tell if Ho or Lo is heating, but guessing is probably LO .
also the waveform distortion at the primary becomes really apparent and im scared that in moments my power supply could turn in a bomb shooting Mosfet pieces all over the place(or the fuse becomes a sharpnel grenade). those are the waveforms that i get at the primary switching node , with no load and at 500w of load

also i removed the snubber and this is the waveform with no snubber
what do you think am i doing wrong? i clamped the Vcc pin of the ic via a 15v 1.3w zener and the supply is at 15v all times, i also tried to reduce the drive resistors from 220ohm to 22ohm but no effect on the output waveform, im placing my bets on the core of the transformer since its the only component that i ordered cheap from china(not that cheap compared to original distributors)
please help me guys !
thanks in advance and sorry for my bad english, also my scope if you notice is older than most of the people on this forum 🧐 sorry for the bad quality of the images
 

uxavier

New member
im using irfp450 mosfets, i also tried with irfp440(much easier to drive)
and i also tried to lower the resistor from 220ohm to 22ohm even 10ohms but there is no difference ....
 
Top